It was good to do a proper running post last week, remembering my trips down the "Lakes in a Day" course and looking forward to this year's race which will hopefully go ahead in October. But for this piece I'm reverting to my current occasional pastime of thinking about the Coronavirus epidemic and what it's doing to us all. I was tempted to head this post with the famous phrase attributed to Benjamin Disraeli - "Lies, Damned Lies and Statistics" but it doesn't quite grasp what I'm after, which is more about how we often latch onto a single piece of factual information and come to rapid conclusions about what it means without considering whether there are other factors in play that might affect the situation. And from there it is only a small step to selecting only the information which may support a stance you already have, and ignoring the rest.
So for this post I'm going to cover a topic the media loves to come back to fairly regularly, that of comparing our situation with other countries. I am told that (a) we have the second worst death rate in the world, and (b) that this is because the policies pursued in the UK have been completely wrong. Everyone else has done better. Why are we not doing the same as Germany, South Korea, New Zealand, Italy, France, Greece, etc? Now this may well be a completely justifiable proposition but I somehow feel it's naive to accept it at face value without thinking for a moment or two about some of the many factors that play into the situation. I'll say at the outset I'm not trying to justify any of the actions that our government has taken. It is not a government I would have chosen, nor one I voted for. It just seems to me that laying the whole blame on policy without even thinking whether other factors might be involved is just lazy.
I'm not going to draw any detailed conclusions on causes from this, I'll just set down some information and speculate on which bits might be relevant. All the figures below are from 10 June when I did the research; they may have changed a bit by now but the overall patterns will still be the same
1. Country "league table" of Coronavirus deaths
Total deaths by country are a meaningless comparison so I have taken deaths per million population as being more informative. I have also omitted tiny countries which do not have a big enough population to be statistically meaningful - for example both San Marino and Andorra are in the top three but I don't think that tells you much about general trends. So here are the 10 worst countries and also 3 from further down the list which are generally considered to be coping pretty well.
Country Deaths/million population
Belgium 831
UK 606
Spain 580
Italy 564
Sweden 475
France 449
Netherlands 353
USA 347
Ireland 344
Switzerland 224
Germany 106
South Korea 5
New Zealand 4
2. Geographical and Physical Factors which may have an influence
I'm not going to clutter this up with lots more lists or tables, they are all readily available for anyone who spends a few minutes on research, but I'll just pick out what I think might be relevant headlines. The comments below came from looking up all the measures concerned for all the countries on the above list.
The virus spreads when people are in close proximity so you would expect the population density of the country to have some influence. The UK's density is 268 people per square kilometer, many times more than USA (33), Sweden (22) and New Zealand (19), but significantly below South Korea (520). The "populous" European countries range from the Netherlands (414) and Belgium (354) through to the relatively sparsely populated France (104) and Spain (92). Germany's density is 232.
Concentration of population into densely populated conurbations will also have an effect. Australia is a classic case of a vast country with almost all its population in cities, but none of the countries on the list above are particularly extreme in this effect - compare the two ends with New Zealand having 1,5m of its 5m population in Auckland with the UK having around 20m of its 65m population in major conurbations, a similar proportion overall.
The number of people visiting from outside the country would also be expected to have an effect, because this increases the number of different people in circulation. This is really only measurable for countries that maintain physical borders. It's easy to compare the 4 million people a year visiting New Zealand (as they have to cross a thousand miles of sea) with the 40 million for the UK (and the rather surprisingly low 77 million for USA), but although countries like Belgium and the Netherlands can say they each have around 20 million visitors a year, those are only people who can be identified as tourists via hotel records, etc; their real exposure to incomers will be far higher due to their completely open land borders, probably at least double these figures.
Any algorithm linking these effects would be very complex, but as a straw man one could argue that on these figures the average Belgian is maybe 200 times more likely to meet a Covid carrier than the average New Zealander, before any countermeasures are taken.
3. Population Characteristics which may have an influence
This is where we move away from factors which may increase or decrease an individual's chance of contracting Covid 19 to those which might affect the outcome if they do get it.
We know now that population age distribution has a significant effect. In the UK 90% of all deaths have been in the over 65 age group, so I looked at proportion of population over 65 in the listed countries. The Europeans are all in a 5% band ranging from 23% in Italy to 18% in the UK . The "non-Europeans" (USA, New Zealand and South Korea) all have over-65 populations of around 15%. This is not a huge band overall but you would expect the countries with the lower percentages to have fewer deaths by some amount due to this alone.This effect is really amplified if you look at developing countries. Nigeria for example has an over-65 population of only 3%, so pro-rata should have only about 15% of the over-65 deaths that the UK has (all other things being equal, which of course they are not).
It's difficult to find useful information on the second major factor that we are told about, that is underlying health conditions. It's easy to find comparison of death rates but from these you can't then tell what proportion is due to the incidence of the disease and what due to the country's success at treating it. Out of interest, for chronic heart diseases, the UK's death rate is double that of South Korea and almost one and a half times that of France, but is still significantly below the rates for New Zealand, Germany and the USA. I wondered if Type 2 Diabetes might be a factor, but rather surprisingly all the countries on the list show as very similar except the USA which is a few percentage points higher. Obesity shows a larger spread, with 36% of the USA population showing as obese compared with 28% in the UK, 20% in Italy and 5% in South Korea.
We are also learning more recently that economic situation has an impact - more poorer people succumb to Covid than richer ones. I looked at the effective incomes of population in my list (this is income corrected for the different cost of living in different countries, so reflects actual purchasing power). The USA comes out as the richest country and South Korea the poorest, with the UK just about the mid point. But this is not the whole story, you could argue that the difference between rich and poor plays a much more significant part than average country income. Again, the USA tops the list with the biggest gap between rich and poor, with the UK in second place. The most egalitarian countries are the Netherlands, Sweden and Belgium. Rather surprisingly South Korea shows up as the third most "unequal" country.
4. Cultural characteristics which may have an influence
This is tricky ground but really cannot be ignored, and is really about how populations react to authority and a "common cause". I'll quote a flippant little example to set the scene. I worked for many years with colleagues from all over Europe and one coffee-break conversation that came up from time to time was the "wet paint" sign. It was generally acknowledged by the relevant representatives of their countries that a "wet paint" sign to a German means "this paint is wet, so I won't touch it", whereas the same sign to a Brit means "I wonder if the paint is dry yet, I'll just check". But the most illustrative point from this discussion was that both the Brits and the Germans, while fully recognising the thumbnail as fairly true to type, each thought that the point of view of the other was something to make a joke about.
I can't really go much further than that, other to observe that there are cultural differences between populations. Some will generally go along with recommendations and rules set up by their governments (one assumes in good faith) to protect them because they believe it is the most effective thing to do. Some will question and look for ways around them. Just as a final comment from my time in gainful employment (long ago now) in an international community. We often found that getting exactly the right solution was less important than getting a solution that everyone believed was right. Solid progress towards 80% always beat continual argument on how to get to 100. So even in responses to something like a pandemic, one would expect united countries who have some faith in their government (regardless of the competency, within limits, of that government) to make better progress than disunited ones.
5. Measuring
All of the above assumes that all countries measure their incidence of infection and death rate in the same way, which of course they don't. We are told that a reliable measure of death rate is the increase in total over the usual rate for the time of year. But this pandemic still has a long way to run yet; there will be deaths already and many more to come I that will be caused by but not directly attributable to Covid. Heart attack and stroke victims that couldn't be got to hospital in time, cancer sufferers not able to get timely treatment, suicides due to economic stress, premature deaths due to deprivation in a drastically depressed economic climate and so on. Different countries will record these in different ways and at different times.
Conclusions
I said I wouldn't make any detailed conclusions, and I won't. But here are just a few observations that are clear to me from the delving into the data that I did:
1. The overall situation is far too complex at present for detailed comparisons to be made between countries. In spite of the politicians declaring that it will be possible in the future, I rather doubt this.
2. It is however possible to see a few broad trends. A group of a half dozen European countries, including the UK, have been hit hard. It seems difficult to pull out whether physical factors or policy have caused the differences between them. Has Spain fared marginally better than the UK because of its draconian lock-down or because it population density is only 30% of the UK's? Should Italy be denigrated for a death rate 93% of the UK's (proclaimed by the world's media as disgraceful) or congratulated for looking after the oldest population in Europe? What factors affected Belgium to give it a death rate more than double its neighbour the Netherlands?
3. Among these countries, and the others of the "top ten", it is difficult as yet to see how their policies have affected their current statistics. Both the USA and Sweden seem to have death rates no worse (and in some cases far better) than the others in spite of patchy or non-existent "lockdown" policies. Is this because lockdowns for the others were ineffective (or came too late) or because Sweden and USA are countries with very low population densities?
4. However, although it is difficult to work out how and why the "poorly performing" countries at the top of the list went wrong, it is clear that some countries did very much better, notably Germany and South Korea (I think comparisons with New Zealand are pointless; they did a good job but with all their geographical advantages would probably have come out relatively well whatever they did). Germany and South Korea are densely populated countries. Germany has more land borders than any other in Europe. They have their share of inherent disease, poverty and inequality. Yet Germany has recorded only one eighth the death rate of Belgium, and South Korea a staggering less than one percent of Belgium's. Their main feature in common is that over the last few generations they have been outstandingly successful in harnessing the collective power of their people to produce two of the most productive economies in the world. Could there be a clue there?
But I'm afraid that my main feeling at the end of this exercise is that it has just confirmed the feeling I had right at the start; that in general the media does not use information to form an opinion but selects isolated items of data to reinforce the opinions they already have. I'll close with an unrelated item. Over the last couple of days the news has been full of the pronouncement, supported by the OECD forecast, that Britain's economy will be "The Worst Hit in Europe" as we try to rebuild after the pandemic. Which is true. I took the trouble to look at all the figures. Yes, the forecast for the UK at an 11.5% downturn looks pretty grim compared with Germany at 6.6%. The figures for France and Italy? 11.4% and 11.3% respectively. Margin for error? Exactly, but let's not mention that.
Tomorrow morning I'm setting out on the Virtual West Highland Way race, so next week I hope to have something a bit more uplifting to write about.
No comments:
Post a Comment